Learning with Kernels on Graphs: DAG-based kernels, data streams and RNA function prediction.

In many application domains data can be naturally represented as graphs. When the application of analytical solutions for a given problem is unfeasible, machine learning techniques could be a viable way to solve the problem. Classical machine learning techniques are defined for data represented in a vectorial form. Recently some of them have been extended to deal directly with structured data. Among those techniques, kernel methods have shown promising results both from the computational complexity and the predictive performance point of view. Kernel methods allow to avoid an explicit mapping in a vectorial form relying on kernel functions, which informally are functions calculating a similarity measure between two entities. However, the definition of good kernels for graphs is a challenging problem because of the difficulty to find a good tradeoff between computational complexity and expressiveness. Another problem we face is learning on data streams, where a potentially unbounded sequence of data is generated by some sources. There are three main contributions in this thesis. The first contribution is the definition of a new family of kernels for graphs based on Directed Acyclic Graphs (DAGs). We analyzed two kernels from this family, achieving state-of-the-art results from both the computational and the classification point of view on real-world datasets. The second contribution consists in making the application of learning algorithms for streams of graphs feasible. Moreover,we defined a principled way for the memory management. The third contribution is the application of machine learning techniques for structured data to non-coding RNA function prediction. In this setting, the secondary structure is thought to carry relevant information. However, existing methods considering the secondary structure have prohibitively high computational complexity. We propose to apply kernel methods on this domain, obtaining state-of-the-art results.

Synthetic strategies of new spin-labelled supramolecular systems

Molecular recognition and self-assembly represent fundamental issues for the construction of supramolecular systems, structures in which the components are held together through non-covalent interactions. The study of host-guest complexes and mechanical interlocked molecules, important examples in this field, is necessary in order to characterize self-assembly processes, achieve more control over the molecular organization and develop sophisticated structures by using properly designed building blocks. The introduction of paramagnetic species, or spin labelling, represents an attractive opportunity that allows their detection and characterization by the Electron Spin Resonance spectroscopy, a valuable technique that provides additional information to those obtained by traditional methods. In this Thesis, recent progresses in the design and the synthesis of new paramagnetic host-guest complexes and rotaxanes characterized by the presence of nitroxide radicals and their investigation by ESR spectroscopy are reported. In Chapter 1 a brief overview of the principal concepts of supramolecular chemistry, the spin labelling approach and the development of ESR methods applied to paramagnetic systems are described. Chapter 2 and 3 are focused on the introduction of radicals in macrocycles as Cucurbiturils and Pillar[n]arenes, due to the interesting binding properties and the potential employment in rotaxanes, in order to investigate their structures and recognition properties. Chapter 4 deals with one of the most studied mechanical interlocked molecules, the bistable [2]rotaxane reported by Stoddart and Heath based on the ciclobis (paraquat-p-phenylene) CBPQT4+, that represents a well known example of molecular switch driven by external stimuli. The spin labelling of analogous architectures allows the monitoring by ESR spectroscopy of the switch mechanism involving the ring compound by tuning the spin exchange interaction. Finally, Chapter 5 contains the experimental procedures used for the synthesis of some of the compounds described in Chapter 2-4.